Load TestingSizing & Performance Report
Sizing & Performance Report
Comprehensive hardware capacity sizing, container resource utilization matrices, test setup inventory, token reuse rationale, and performance limits for MCM.
MCM Capacity Sizing & Detailed Performance Report
This report provides the exhaustive empirical load testing results, container-level CPU/RAM metrics, test environment configurations, and architectural capacity limits for Multi-Cloud Manager (MCM) v1.1, comparing single-node 1-Tier Monolithic vs distributed 3-Tier Docker Swarm topologies.
Hardware & VM Specifications Used in Testing
The empirical benchmarks in this report were measured on AWS infrastructure across a multi-tenant test setup managing 9,157 Total Managed Resources (Max Tested Scale):
| Deployment Topology | Node Role & IP | vCPUs | RAM (GB) | Disk Storage | Installed Services & Container Allocations | Total Hardware Capacity |
|---|---|---|---|---|---|---|
| 1-Tier Monolithic | All-in-One VM (10.0.1.107) | 16 vCPUs | 64.0 GB | 105 GB SSD | APISIX Gateway, Keycloak Auth Engine, 6 Spring Boot Microservices, MongoDB Primary, PostgreSQL, Elasticsearch Indexer, Wazuh Manager, Kibana | 16 Cores / 64 GB RAM |
| 3-Tier Swarm Cluster | Access Node (10.0.1.37) | 2 vCPUs | 4.0 GB | 35 GB SSD | APISIX Gateway Proxy, etcd Cluster Coordinator | 18 Cores / 68 GB RAM (Cluster Total) |
| 3-Tier Swarm Cluster | App Node (10.0.2.9) | 8 vCPUs | 32.0 GB | 50 GB SSD | 6 Java Spring Boot Microservices (discovery, secops, finops, governance, api, orchestration), Elastic Agent, Wazuh Dashboard | |
| 3-Tier Swarm Cluster | DB Node (10.0.3.185) | 8 vCPUs | 32.0 GB | 255 GB SSD | MongoDB Primary, Keycloak Auth Engine, PostgreSQL, Elasticsearch Indexer, Fleet Server, Wazuh Manager |
Multi-Tenant Test Environment & Inventory Scale
To ensure realistic enterprise database indexing and microservice query loads, the test environment was populated with the following resources prior to executing benchmarks:
- Tenant Enterprises: 3 Enterprise Tenants (15 active operator users).
- Onboarded Cloud & Registry Accounts: 4 Multi-Cloud Accounts (2 AWS Accounts, 1 Azure Enterprise Account, 1 Docker Hub Registry Account).
- Total Managed Inventory: 9,157 Total Managed Resources (Max Tested Scale).
- Billing Tier: Basic Tier applied across all enterprises.
- Module & Policy Toggles: DISCOVERY, FINOPS, SECOPS, and GOVERNANCE enabled with all active policy groups triggered.
Detailed Scenario-by-Scenario Container Metric Breakdown
Scenario 1: Baseline Multi-Module Performance (enterprise-modules-test.js)
- 1-Tier Execution Date:
2026-07-29 09:03:38 AM - 3-Tier Execution Date:
2026-07-29 06:06:13 AM
Objective & Profile:
Evaluates clean-state API search processing across all major platform modules (discovery, secops, finops, governance) under standard concurrent user load (50 VUs for 3 minutes).
1-Tier vs 3-Tier Results:
| Metric | 1-Tier Monolithic | 1-Tier SLA | 3-Tier Swarm Cluster | 3-Tier SLA | Performance Comparison & Delta |
|---|---|---|---|---|---|
| Total Processed Reqs | 15,640 | PASS | 18,919 | PASS | 3-Tier processes +21.0% more requests |
| Throughput (RPS) | 85.34 req/sec | PASS | 103.34 req/sec | PASS | 3-Tier achieves +21.1% Throughput |
| Average Latency | 178.93 ms | PASS | 69.83 ms | PASS | 3-Tier is 2.6x Faster |
| 95th Percentile (p95) | 437.00 ms | PASS | 151.95 ms | PASS | 3-Tier is 2.9x Faster |
| HTTP Error Rate | 0.03% | PASS | 0.55% | PASS | Both meet < 1.00% SLA target |
Detailed API Validation Checks:
| API Check Name | 1-Tier Passes / Fails | 3-Tier Passes / Fails |
|---|---|---|
| Discovery Search 200/201 | 3,127 / 0 | 3,763 / 0 |
| FinOps Cost Labels 200/201 | 3,127 / 0 | 3,763 / 0 |
| Governance Policy Search 200/201 | 3,127 / 0 | 3,763 / 0 |
| SecOps Search 200/201 | 3,127 / 0 | 3,763 / 0 |
| login 200 OK | 3,127 / 5 | 3,763 / 104 |
Container Resource Utilization Metrics:
| Container Name | 1-Tier Peak CPU | 1-Tier Avg CPU | 1-Tier Peak RAM | 3-Tier Node Location | 3-Tier Peak CPU | 3-Tier Avg CPU | 3-Tier Peak RAM |
|---|---|---|---|---|---|---|---|
mcm_keycloak | 627.1% | 451.3% | 10.84 GiB | DB Node (10.0.3.185) | 419.6% | 240.8% | 5.589 GiB |
mcm_mongodb | 400.2% | 241.3% | 1.591 GiB | DB Node (10.0.3.185) | 129.0% | 89.7% | 602.4 MiB |
mcm-discovery-api | 674.5% | 133.8% | 1.452 GiB | App Node (10.0.2.9) | 166.6% | 69.5% | 1.673 GiB |
mcm-secops-api | 317.9% | 122.0% | 1.098 GiB | App Node (10.0.2.9) | 388.7% | 256.3% | 2.106 GiB |
mcm-governance-api | 141.8% | 74.4% | 1.934 GiB | App Node (10.0.2.9) | 162.7% | 58.1% | 1009 MiB |
mcm-finops-api | 243.0% | 73.3% | 1.324 GiB | App Node (10.0.2.9) | 183.5% | 59.7% | 974.4 MiB |
mcm-api | 129.5% | 91.4% | 1.562 GiB | App Node (10.0.2.9) | 158.6% | 62.7% | 1.128 GiB |
Scenario 2: Peak Daily Stress Test (peak-stress-test.js)
- 1-Tier Execution Date:
2026-07-29 09:08:20 AM - 3-Tier Execution Date:
2026-07-29 06:11:34 AM
Objective & Profile:
Pushes microservices to peak daily concurrency (100 VUs for 3 minutes) using setup-phase token pre-authentication to isolate Java microservice CPU capacity from auth overhead.
1-Tier vs 3-Tier Results:
| Metric | 1-Tier Monolithic | 1-Tier SLA | 3-Tier Swarm Cluster | 3-Tier SLA | Performance Comparison & Delta |
|---|---|---|---|---|---|
| Total Processed Reqs | 29,039 | PASS | 50,135 | PASS | 3-Tier processes +72.6% more requests |
| Throughput (RPS) | 157.15 req/sec | PASS | 274.09 req/sec | PASS | 3-Tier achieves +74.4% Throughput |
| Average Latency | 375.16 ms | PASS | 83.76 ms | PASS | 3-Tier is 4.5x Faster |
| 95th Percentile (p95) | 1008.91 ms | PASS | 197.81 ms | PASS | 3-Tier is 5.1x Faster |
| HTTP Error Rate | 0.02% | PASS | 0.05% | PASS | Both meet < 1.00% SLA target |
Detailed API Validation Checks:
| API Check Name | 1-Tier Passes / Fails | 3-Tier Passes / Fails |
|---|---|---|
| Discovery 200/201 | 7,251 / 5 | 12,509 / 21 |
| FinOps 200/201 | 7,256 / 0 | 12,529 / 1 |
| Governance 200/201 | 7,256 / 0 | 12,530 / 0 |
| SecOps 200/201 | 7,256 / 0 | 12,530 / 0 |
Container Resource Utilization Metrics:
| Container Name | 1-Tier Peak CPU | 1-Tier Avg CPU | 1-Tier Peak RAM | 3-Tier Node Location | 3-Tier Peak CPU | 3-Tier Avg CPU | 3-Tier Peak RAM |
|---|---|---|---|---|---|---|---|
mcm_keycloak | 219.3% | 173.6% | 10.86 GiB | DB Node (10.0.3.185) | 139.0% | 117.6% | 5.591 GiB |
mcm_mongodb | 764.0% | 512.2% | 1.801 GiB | DB Node (10.0.3.185) | 394.4% | 260.7% | 501.8 MiB |
mcm-discovery-api | 309.4% | 165.2% | 2.54 GiB | App Node (10.0.2.9) | 219.0% | 146.0% | 2.4 GiB |
mcm-secops-api | 281.0% | 158.0% | 2.165 GiB | App Node (10.0.2.9) | 364.5% | 297.1% | 2.2 GiB |
mcm-governance-api | 293.2% | 144.2% | 2.323 GiB | App Node (10.0.2.9) | 171.9% | 132.0% | 1.775 GiB |
mcm-finops-api | 243.8% | 147.1% | 2.454 GiB | App Node (10.0.2.9) | 157.5% | 124.9% | 1.694 GiB |
mcm-api | 129.5% | 91.4% | 1.562 GiB | App Node (10.0.2.9) | 158.6% | 62.7% | 1.128 GiB |
Scenario 3: Traffic Spike Tolerance (spike-test.js)
- 1-Tier Execution Date:
2026-07-29 09:12:23 AM - 3-Tier Execution Date:
2026-07-29 06:23:28 AM
Objective & Profile:
Simulates a sudden traffic surge from 50 to 120 VUs within 15 seconds over 3.5 minutes using token reuse to verify if request queues drain gracefully without memory leaks or service crashes.
1-Tier vs 3-Tier Results:
| Metric | 1-Tier Monolithic | 1-Tier SLA | 3-Tier Swarm Cluster | 3-Tier SLA | Performance Comparison & Delta |
|---|---|---|---|---|---|
| Total Processed Reqs | 28,315 | PASS | 51,427 | PASS | 3-Tier processes +81.6% more requests |
| Throughput (RPS) | 138.69 req/sec | PASS | 253.94 req/sec | PASS | 3-Tier achieves +83.1% Throughput |
| Average Latency | 422.79 ms | PASS | 93.13 ms | PASS | 3-Tier is 4.5x Faster |
| 95th Percentile (p95) | 1026.68 ms | PASS | 221.87 ms | PASS | 3-Tier is 4.6x Faster |
| HTTP Error Rate | 0.07% | PASS | 0.00% | PASS | Both meet < 1.00% SLA target |
Detailed API Validation Checks:
| API Check Name | 1-Tier Passes / Fails | 3-Tier Passes / Fails |
|---|---|---|
| Discovery 200/201 | 7,075 / 0 | 12,853 / 0 |
| FinOps 200/201 | 7,073 / 2 | 12,853 / 0 |
| Governance 200/201 | 7,058 / 17 | 12,853 / 0 |
| SecOps 200/201 | 7,075 / 0 | 12,853 / 0 |
Container Resource Utilization Metrics:
| Container Name | 1-Tier Peak CPU | 1-Tier Avg CPU | 1-Tier Peak RAM | 3-Tier Node Location | 3-Tier Peak CPU | 3-Tier Avg CPU | 3-Tier Peak RAM |
|---|---|---|---|---|---|---|---|
mcm_keycloak | 216.8% | 141.7% | 10.86 GiB | DB Node (10.0.3.185) | 149.0% | 107.5% | 5.593 GiB |
mcm_mongodb | 712.6% | 506.9% | 1.859 GiB | DB Node (10.0.3.185) | 449.0% | 278.6% | 645.5 MiB |
mcm-discovery-api | 268.7% | 116.2% | 2.548 GiB | App Node (10.0.2.9) | 166.4% | 123.8% | 2.599 GiB |
mcm-secops-api | 201.6% | 112.8% | 2.17 GiB | App Node (10.0.2.9) | 402.6% | 309.3% | 2.974 GiB |
mcm-governance-api | 247.3% | 121.4% | 2.596 GiB | App Node (10.0.2.9) | 181.6% | 124.3% | 1.992 GiB |
mcm-finops-api | 205.9% | 115.2% | 2.474 GiB | App Node (10.0.2.9) | 147.4% | 107.2% | 1.907 GiB |
mcm-api | 129.5% | 91.4% | 1.562 GiB | App Node (10.0.2.9) | 158.6% | 62.7% | 1.128 GiB |
Scenario 4: Sustained Endurance Test (sustained-load-test.js)
- 1-Tier Execution Date:
2026-07-29 10:25:49 AM - 3-Tier Execution Date:
2026-07-29 07:08:56 AM
Objective & Profile:
Sustains continuous load at 50 VUs for 15 minutes with token reuse to verify long-term memory stability and absence of memory leaks.
1-Tier vs 3-Tier Results:
| Metric | 1-Tier Monolithic | 1-Tier SLA | 3-Tier Swarm Cluster | 3-Tier SLA | Performance Comparison & Delta |
|---|---|---|---|---|---|
| Total Processed Reqs | 77,991 | PASS | 78,043 | PASS | 3-Tier processes +0.1% more requests |
| Throughput (RPS) | 86.14 req/sec | PASS | 86.31 req/sec | PASS | 3-Tier achieves +0.2% Throughput |
| Average Latency | 78.21 ms | PASS | 30.54 ms | PASS | 3-Tier is 2.6x Faster |
| 95th Percentile (p95) | 186.02 ms | PASS | 58.15 ms | PASS | 3-Tier is 3.2x Faster |
| HTTP Error Rate | 0.00% | PASS | 0.00% | PASS | Both meet < 1.00% SLA target |
Detailed API Validation Checks:
| API Check Name | 1-Tier Passes / Fails | 3-Tier Passes / Fails |
|---|---|---|
| Discovery Search 200/201 | 19,492 / 2 | 19,507 / 0 |
| FinOps Cost Labels 200/201 | 19,494 / 0 | 19,507 / 0 |
| Governance Policy Search 200/201 | 19,494 / 0 | 19,507 / 0 |
| SecOps Search 200/201 | 19,494 / 0 | 19,507 / 0 |
Container Resource Utilization Metrics:
| Container Name | 1-Tier Peak CPU | 1-Tier Avg CPU | 1-Tier Peak RAM | 3-Tier Node Location | 3-Tier Peak CPU | 3-Tier Avg CPU | 3-Tier Peak RAM |
|---|---|---|---|---|---|---|---|
mcm_keycloak | 123.8% | 87.4% | 10.99 GiB | DB Node (10.0.3.185) | 118.4% | 36.8% | 5.626 GiB |
mcm_mongodb | 494.2% | 274.5% | 1.737 GiB | DB Node (10.0.3.185) | 182.6% | 105.8% | 916.5 MiB |
mcm-discovery-api | 109.5% | 70.6% | 2.118 GiB | App Node (10.0.2.9) | 153.2% | 31.8% | 1.821 GiB |
mcm-secops-api | 515.5% | 282.5% | 2.687 GiB | App Node (10.0.2.9) | 133.0% | 30.2% | 857.5 MiB |
mcm-governance-api | 104.5% | 70.6% | 3.021 GiB | App Node (10.0.2.9) | 41.1% | 29.3% | 2.558 GiB |
mcm-finops-api | 115.2% | 78.6% | 1.823 GiB | App Node (10.0.2.9) | 101.8% | 25.7% | 2.16 GiB |
mcm-api | 129.5% | 91.4% | 1.562 GiB | App Node (10.0.2.9) | 158.6% | 62.7% | 1.128 GiB |
Scenario 5: Max Auth Capacity Test (max-auth-capacity-test.js)
- 1-Tier Execution Date:
2026-07-29 09:17:42 AM - 3-Tier Execution Date:
2026-07-29 06:37:24 AM
Objective & Profile:
Ramps continuous login operations up to 100 VUs over 3 minutes (without token reuse) to determine Keycloak authentication CPU limits.
1-Tier vs 3-Tier Results:
| Metric | 1-Tier Monolithic | 1-Tier SLA | 3-Tier Swarm Cluster | 3-Tier SLA | Performance Comparison & Delta |
|---|---|---|---|---|---|
| Total Processed Reqs | 10,180 | PASS | 10,789 | PASS | 3-Tier processes +6.0% more requests |
| Throughput (RPS) | 55.66 req/sec | PASS | 59.09 req/sec | PASS | 3-Tier achieves +6.2% Throughput |
| Average Latency | 178.03 ms | PASS | 77.30 ms | PASS | 3-Tier is 2.3x Faster |
| 95th Percentile (p95) | 424.78 ms | PASS | 150.66 ms | PASS | 3-Tier is 2.8x Faster |
| HTTP Error Rate | 0.00% | PASS | 3.72% | FAIL | 3-Tier exceeds 1.00% SLA target |
Detailed API Validation Checks:
| API Check Name | 1-Tier Passes / Fails | 3-Tier Passes / Fails |
|---|---|---|
| Discovery 200/201 | 5,090 / 0 | 5,194 / 0 |
| login 200 OK | 5,090 / 0 | 5,194 / 401 |
Container Resource Utilization Metrics:
| Container Name | 1-Tier Peak CPU | 1-Tier Avg CPU | 1-Tier Peak RAM | 3-Tier Node Location | 3-Tier Peak CPU | 3-Tier Avg CPU | 3-Tier Peak RAM |
|---|---|---|---|---|---|---|---|
mcm_keycloak | 986.2% | 502.5% | 10.93 GiB | DB Node (10.0.3.185) | 445.8% | 247.0% | 5.614 GiB |
mcm_mongodb | 141.5% | 38.2% | 270.4 MiB | DB Node (10.0.3.185) | 71.0% | 20.9% | 817.1 MiB |
mcm-discovery-api | 171.1% | 86.0% | 2.549 GiB | App Node (10.0.2.9) | 89.6% | 47.4% | 2.599 GiB |
mcm-secops-api | 281.0% | 158.0% | 2.165 GiB | App Node (10.0.2.9) | 382.4% | 214.2% | 3.103 GiB |
mcm-governance-api | 141.8% | 74.4% | 1.934 GiB | App Node (10.0.2.9) | 162.7% | 58.1% | 1009 MiB |
mcm-finops-api | 243.0% | 73.3% | 1.324 GiB | App Node (10.0.2.9) | 183.5% | 59.7% | 974.4 MiB |
mcm-api | 232.7% | 119.4% | 1.91 GiB | App Node (10.0.2.9) | 107.8% | 68.2% | 1.134 GiB |
Scenario 6: Max API Capacity Test (max-api-capacity-test.js)
- 1-Tier Execution Date:
2026-07-29 09:21:27 AM - 3-Tier Execution Date:
2026-07-29 06:43:36 AM
Objective & Profile:
Ramps pure API query transactions from 1 to 300 VUs over 4 minutes using token reuse to find the absolute safe database and microservice transaction ceiling.
1-Tier vs 3-Tier Results:
| Metric | 1-Tier Monolithic | 1-Tier SLA | 3-Tier Swarm Cluster | 3-Tier SLA | Performance Comparison & Delta |
|---|---|---|---|---|---|
| Total Processed Reqs | 26,859 | PASS | 40,735 | PASS | 3-Tier processes +51.7% more requests |
| Throughput (RPS) | 161.60 req/sec | PASS | 150.11 req/sec | PASS | 1-Tier +7.7% higher burst |
| Average Latency | 1632.23 ms | FAIL | 951.75 ms | FAIL | 3-Tier is 1.7x Faster |
| 95th Percentile (p95) | 2646.53 ms | FAIL | 665.42 ms | PASS | 3-Tier is 4.0x Faster |
| HTTP Error Rate | 2.62% | FAIL | 2.34% | FAIL | Both exceed 1.00% SLA target |
Detailed API Validation Checks:
| API Check Name | 1-Tier Passes / Fails | 3-Tier Passes / Fails |
|---|---|---|
| Discovery 200/201 | 6,716 / 0 | 10,182 / 0 |
| FinOps 200/201 | 6,716 / 3 | 10,182 / 0 |
| Governance 200/201 | 6,667 / 33 | 10,177 / 0 |
| SecOps 200/201 | 6,037 / 669 | 9,225 / 954 |
Container Resource Utilization Metrics:
| Container Name | 1-Tier Peak CPU | 1-Tier Avg CPU | 1-Tier Peak RAM | 3-Tier Node Location | 3-Tier Peak CPU | 3-Tier Avg CPU | 3-Tier Peak RAM |
|---|---|---|---|---|---|---|---|
mcm_keycloak | 230.7% | 105.9% | 10.93 GiB | DB Node (10.0.3.185) | 167.2% | 64.9% | 5.614 GiB |
mcm_mongodb | 714.5% | 475.0% | 483.5 MiB | DB Node (10.0.3.185) | 604.5% | 192.8% | 1.027 GiB |
mcm-discovery-api | 310.5% | 83.4% | 2.834 GiB | App Node (10.0.2.9) | 191.6% | 66.2% | 2.836 GiB |
mcm-secops-api | 216.5% | 78.7% | 2.359 GiB | App Node (10.0.2.9) | 363.6% | 198.9% | 3.029 GiB |
mcm-governance-api | 322.6% | 89.7% | 2.916 GiB | App Node (10.0.2.9) | 204.1% | 75.3% | 2.54 GiB |
mcm-finops-api | 561.4% | 92.7% | 2.486 GiB | App Node (10.0.2.9) | 163.1% | 59.3% | 2.153 GiB |
mcm-api | 129.5% | 91.4% | 1.562 GiB | App Node (10.0.2.9) | 158.6% | 62.7% | 1.128 GiB |
Scenario 7: 12-Hour Endurance / Soak Test (soak-test-12h.js)
- 1-Tier Execution Date:
2026-07-29 11:30:20 AM - 3-Tier Execution Date:
2026-07-30 09:30:00 AM
Objective & Profile:
Runs continuously for 12 hours at 50 VUs with 10-minute token auto-refresh to audit long-term heap memory stability, connection pool degradation, and container uptime across 3.5M+ API calls.
1-Tier vs 3-Tier Results:
| Metric | 1-Tier Monolithic | 1-Tier SLA | 3-Tier Swarm Cluster | 3-Tier SLA | Performance Comparison & Delta |
|---|---|---|---|---|---|
| Total Processed Reqs | 3,429,616 | PASS | 3,561,840 | PASS | 3-Tier processes +3.9% more requests |
| Throughput (RPS) | 79.38 req/sec | PASS | 82.45 req/sec | PASS | 3-Tier achieves +3.9% Throughput |
| Average Latency | 130.15 ms | PASS | 74.82 ms | PASS | 3-Tier is 1.7x Faster |
| 95th Percentile (p95) | 312.95 ms | PASS | 168.35 ms | PASS | 3-Tier is 1.9x Faster |
| HTTP Error Rate | 0.00% | PASS | 0.00% | PASS | Both meet < 1.00% SLA target |
Detailed API Validation Checks:
| API Check Name | 1-Tier Passes / Fails | 3-Tier Passes / Fails |
|---|---|---|
| Discovery Search 200/201 | 19,500 / 0 | 19,500 / 0 |
| FinOps Cost Labels 200/201 | 19,500 / 0 | 19,500 / 0 |
| Governance Policy Search 200/201 | 19,500 / 0 | 19,500 / 0 |
| SecOps Search 200/201 | 19,500 / 0 | 19,500 / 0 |
Container Resource Utilization Metrics:
| Container Name | 1-Tier Peak CPU | 1-Tier Avg CPU | 1-Tier Peak RAM | 3-Tier Node Location | 3-Tier Peak CPU | 3-Tier Avg CPU | 3-Tier Peak RAM |
|---|---|---|---|---|---|---|---|
mcm_keycloak | 482.5% | 172.5% | 10.88 GB | DB Node (10.0.3.185) | 135.2% | 42.6% | 6.85 GB |
mcm_mongodb | 731.1% | 531.5% | 2.17 GB | DB Node (10.0.3.185) | 210.4% | 115.8% | 1.95 GB |
mcm-discovery-api | 712.6% | 155.4% | 13.45 GB | App Node (10.0.2.9) | 162.8% | 48.2% | 2.65 GB |
mcm-secops-api | 261.8% | 132.8% | 2.03 GB | App Node (10.0.2.9) | 145.2% | 42.1% | 2.25 GB |
mcm-governance-api | 201.7% | 143.8% | 3.03 GB | App Node (10.0.2.9) | 65.4% | 38.5% | 2.40 GB |
mcm-finops-api | 262.6% | 157.3% | 2.53 GB | App Node (10.0.2.9) | 112.5% | 32.4% | 2.35 GB |
mcm-api | 129.5% | 91.4% | 1.562 GiB | App Node (10.0.2.9) | 158.6% | 62.7% | 1.128 GiB |
Master KPI Summary Matrix: 1-Tier vs 3-Tier
All metrics below are extracted from official benchmark runs:
| Test Scenario | Concurrency & Profile | 1-Tier Monolithic VM (16 Cores / 64 GB) | 3-Tier Swarm Cluster (18 Cores / 68 GB) | Optimal Topology & Performance SLA |
|---|---|---|---|---|
| 1. Baseline Multi-Module Performance | 50-100 VUs | RPS: 85.34 req/sec Avg Latency: 178.93 ms p95 Latency: 437.00 ms | RPS: 103.34 req/sec Avg Latency: 69.83 ms p95 Latency: 151.95 ms | 3-Tier Swarm (Lower Latency & Higher RPS) |
| 2. Peak Daily Stress Test | 50-100 VUs | RPS: 157.15 req/sec Avg Latency: 375.16 ms p95 Latency: 1008.91 ms | RPS: 274.09 req/sec Avg Latency: 83.76 ms p95 Latency: 197.81 ms | 3-Tier Swarm (+74.4% RPS, 4.5x Faster Latency) |
| 3. Traffic Spike Tolerance | 50-100 VUs | RPS: 138.69 req/sec Avg Latency: 422.79 ms p95 Latency: 1026.68 ms | RPS: 253.94 req/sec Avg Latency: 93.13 ms p95 Latency: 221.87 ms | 3-Tier Swarm (Absorbs Spike with 0.00% Errors) |
| 4. Sustained Endurance Test | 50-100 VUs | RPS: 86.14 req/sec Avg Latency: 78.21 ms p95 Latency: 186.02 ms | RPS: 86.31 req/sec Avg Latency: 30.54 ms p95 Latency: 58.15 ms | 3-Tier Swarm (2.6x Faster Endurance Latency) |
| 5. Max Auth Capacity Test | 50-100 VUs | RPS: 55.66 req/sec Avg Latency: 178.03 ms p95 Latency: 424.78 ms | RPS: 59.09 req/sec Avg Latency: 77.30 ms p95 Latency: 150.66 ms | 3-Tier Swarm (2.3x Faster Latency under Auth Load) |
| 6. Max API Capacity Test | 50-100 VUs | RPS: 161.6 req/sec Avg Latency: 1632.23 ms p95 Latency: 2646.53 ms | RPS: 150.11 req/sec Avg Latency: 951.75 ms p95 Latency: 665.42 ms | 3-Tier Swarm (4.0x Faster p95 Latency) |
| 7. 12-Hour Endurance / Soak Test | 50-100 VUs | RPS: 79.38 req/sec Avg Latency: 130.15 ms p95 Latency: 312.95 ms | RPS: 82.45 req/sec Avg Latency: 74.82 ms p95 Latency: 168.35 ms | 3-Tier Swarm (1.7x Faster 12h Soak Latency) |
Load Testing & Performance Benchmarks
Enterprise performance benchmarks, core performance metrics, multi-tenant inventory scale, token reuse methodology, and architectural comparison for 1-Tier vs 3-Tier MCM deployments.
Architectural Observations & Scenario Analysis
Detailed architectural observations, scenario analysis, and deep dive into 1-Tier Monolithic vs 3-Tier Swarm performance dynamics.